IP Library Granted Patent US 8,935,463
Granted Patent B1
US 8,935,463 · App. 14/147,462 · Granted Jan 13, 2015

Compute engine in a smart SSD exploiting locality of data

View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 8,935,463
App. No.
14/147,462
Granted
Jan 13, 2015
Kind
B1
Abstract

An embodiment of the invention includes a storage subsystem having a storage central processing unit (SCPU) operable to receive and send a command to a host, the command requiring data computation, a compute engine coupled to the SCPU, and a bank of memory devices coupled to the SCPU and the compute engine and configured to store data required by the commands, wherein the SCPU or the compute engine are operable to perform computation of the data.

Claims (86)

1. A solid state disk (SSD) comprising:

a storage central processing unit (SCPU) operable to receive from and send commands to a host, located externally to the SSD, the SCPU responsive to a command requiring searching of data;

a single compute engine coupled to the SCPU; and

a bank of memory devices coupled to the SCPU and the single compute engine, through an interface, the bank of memory devices configured to store data associated with the command, the single compute engine being separated from the stored data in the bank of memory devices only by the interface, the single compute engine operable to search the stored data entirely within the SSD by exploiting the close locality of the data being searched before the transformed data is sent to the host and to avoid the need for any data movement to and from the host during the search.

2. The solid state disk (SSD), as recited in claim 1 , further including a memory coupled to the SCPU configured to store the data received from the host when the single compute engine is to perform the search.

3. The solid state disk (SSD), as recited in claim 1 , further including a host interface coupled to the SCPU and operable to receive from and send commands to the host.

4. The solid state disk (SSD), as recited in claim 3 , wherein the SSD is operable to avoid moving additional data received from the host, by the host interface, to the bank of memory devices during the transformation thereby reducing power consumption.

5. The solid state disk (SSD), as recited in claim 1 , wherein the memory devices of the bank of memory devices are non-volatile memory.

6. The solid state disk (SSD), as recited in claim 1 , wherein the memory devices of the bank of memory devices are volatile memory.

7. The solid state disk (SSD), as recited in claim 1 , further including a memory controller coupled to the bank of memory devices and operable to control the bank of memory devices.

8. The SSD, as recited in claim 7 , wherein the interface is a PCI Express (PCIe) bus or a processor interconnect bus.

9. The solid state disk (SSD), as recited in claim 1 , wherein the SSD causes reduced latency by avoiding movement of the data across multiple interfaces.

10. The SSD, as recited in claim 1 , wherein the SSD is a server.

11. The SSD, as recited in claim 1 , wherein the host interface is coupled to the host through a host interface bus.

12. The SSD, as recited in claim 11 , wherein the host interface bus is PCI Express (PCIe), Ethernet, fiber channel, infiniband, or Serial ATA (SATA).

13. A solid state disk (SSD) comprising:

a storage central processing unit (SCPU) operable to receive from and send events to a host, located externally to the SSD, the SCPU responsive to an event requiring data transformation;

a single compute engine coupled to the SCPU; and

a bank of memory devices coupled to the SCPU and the single compute engine, through an interface, the bank of memory devices configured to store data associated with the event, the single compute engine being separated from the stored data in the bank of memory devices only by the interface, the single compute engine operable to search the stored data entirely within the SSD by exploiting the close locality of the data being transformed before the transformed data is sent to the host, in response to the command, and to avoid the need for data movement to and from the host during the transformation.

14. The SSD, as recited in claim 13 , wherein the SCPU is operable to determine whether the SCPU or the single compute engine perform the search of the data.

15. The SSD, as recited in claim 13 , further including a memory coupled to the SCPU configured to store the data when the SCPU or the single compute engine is to search the data.

16. The SSD, as recited in claim 13 , wherein the host interface is coupled to the host through a host interface bus.

17. The SSD, as recited in claim 16 , wherein the host interface bus is PCI Express (PCIe), Ethernet, fiber channel, infiniband, or Serial ATA (SATA).

18. The SSD, as recited in claim 13 , wherein the memory devices of the bank of memory devices are non-volatile memory.

19. The SSD, as recited in claim 13 , wherein the memory devices of the bank of memory devices are volatile memory.

20. The SSD, as recited in claim 13 , wherein the storage subsystem is a solid storage disk.

21. The SSD, as recited in claim 13 , wherein the storage subsystem is a server.

22. A method of managing data in a solid state disk (SSD) comprising:

receiving a command and accompanying data from a host through an interface, the host being located externally to the SSD, by a storage central processing unit (SCPU), the command requiring transformation of the data;

storing the data to be transformed in a bank of memory devices coupled to the SCPU;

determining a single compute engine, coupled to the SCPU and the bank of memory devices, is to perform transformation of the data, the single compute engine being separated from the stored data in the bank of memory devices only by the interface; and

upon the determination, the single compute engine being operable to search the stored data entirely within the SSD by exploiting the close locality of the data being search to the single compute engine, before the transformed data is sent to the host and to avoid the need for data movement to and from the host during the search.

23. The method of managing data, as recited in claim 22 , further including moving the data from the bank of memory devices to a memory coupled to the SCPU when the SCPU is to search the data.

24. A solid state disk (SSD) comprising:

a storage central processing unit (SCPU) operable to receive from and send events to a host, located externally to the SSD, the SCPU responsive to an event requiring data fast fourier transformation;

a single compute engine coupled to the SCPU; and

a bank of memory devices coupled to the SCPU and the single compute engine, through an interface, the bank of memory devices configured to store data associated with the event, the single compute engine being separated from the stored data in the bank of memory devices only by the interface, the single compute engine operable to perform fast fourier transformation on the stored data entirely within the SSD by exploiting the close locality of the data being transformed before the transformed data is sent to the host, in response to the command, and to avoid the need for data movement to and from the host during the transformation.

25. A solid state disk (SSD) comprising:

a storage central processing unit (SCPU) operable to receive from and send commands to a host located externally to the SSD, the SCPU responsive to a command requiring a search for data;

a single compute engine coupled to the SCPU; and

a bank of memory devices coupled to the SCPU and the single compute engine and configured to store the data, the single compute engine being separated from the bank of memory devices only through an interface, the SCPU operable to search for the data entirely within the SSD, before a result of the search is sent to the host such as to avoid the need for any data movement to and from the host during the search.

26. A solid state disk (SSD) comprising:

a storage central processing unit (SCPU) operable to receive from and send commands to a host, located externally to the SSD, the SCPU responsive to a command requiring data transformation;

a single compute engine coupled to the SCPU; and

a bank of memory devices coupled to the SCPU and the single compute engine, through an interface, the bank of memory devices configured to store data associated with the command, the single compute engine being separated from the stored data in the bank of memory devices only by the interface, the single compute engine operable to transform the stored data entirely within the SSD by exploiting the close locality of the data being transformed before the transformed data is sent to the host and to avoid the need for any data movement to and from the host during the transformation, wherein the SCPU is operable to determine whether the SCPU or the single compute engine perform the transformation of the data.

27. The solid state disk (SSD), as recited in claim 26 , further including a memory coupled to the SCPU configured to store the data received from the host when the single compute engine is to perform the transformation.

28. The solid state disk (SSD), as recited in claim 26 , further including a host interface coupled to the SCPU and operable to receive from and send commands to the host.

29. The solid state disk (SSD), as recited in claim 28 , wherein the SSD is operable to avoid moving additional data received from the host, by the host interface, to the bank of memory devices during the transformation thereby reducing power consumption.

30. The solid state disk (SSD), as recited in claim 26 , wherein the memory devices of the bank of memory devices are non-volatile memory.

31. The solid state disk (SSD), as recited in claim 26 , wherein the memory devices of the bank of memory devices are volatile memory.

32. The solid state disk (SSD), as recited in claim 26 , further including a memory controller being coupled to the bank of memory devices and operable to control the bank of memory devices.

33. The solid state disk (SSD), as recited in claim 26 , wherein the SSD causes reduced latency by avoiding movement of the data across multiple interfaces.

34. The SSD, as recited in claim 26 , wherein the transformation is fast fourier transform (FFT).

35. The SSD, as recited in claim 26 , wherein the SSD is a server.

36. The SSD, as recited in claim 26 , wherein the interface is a PCI Express (PCIe) bus or a processor interconnect bus.

37. The SSD, as recited in claim 26 , wherein the host interface is coupled to the host through a host interface bus.

38. The SSD, as recited in claim 37 , wherein the host interface bus is PCI Express (PCIe), Ethernet, fiber channel, infiniband, or Serial ATA (SATA).

39. The SSD, as recited in claim 26 , wherein the memory devices of the bank of memory devices are non-volatile memory.

40. The SSD, as recited in claim 26 , wherein the memory devices of the bank of memory devices are volatile memory.

41. The SSD, as recited in claim 26 , wherein the storage subsystem is a solid storage disk.

42. The SSD, as recited in claim 26 , wherein the transformation is fast fourier transform (FFT).

43. The SSD, as recited in claim 26 , wherein the storage subsystem is a server.

44. The SSD, as recited in claim 26 , wherein transforming the data is searching the data.

45. A solid state disk (SSD) comprising:

a storage central processing unit (SCPU) operable to receive from and send events to a host, located externally to the SSD, the SCPU responsive to an event requiring data transformation;

a single compute engine coupled to the SCPU; and

a bank of memory devices coupled to the SCPU and the compute engine, through an interface, the bank of memory devices configured to store data associated with the event, the single compute engine being separated from the stored data in the bank of memory devices only by the interface, the single compute engine operable to transform the stored data entirely within the SSD by exploiting the close locality of the data being transformed before the transformed data is sent to the host, in response to the command, and to avoid the need for data movement to and from the host during the transformation,

wherein the SCPU is operable to determine whether the SCPU or the single compute engine perform the transformation of the data.

46. The SSD, as recited in claim 45 , wherein the SCPU is operable to determine whether the SCPU or the single compute engine perform the transformation of the data.

47. The SSD, as recited in claim 46 , further including a memory coupled to the SCPU configured to store the data when the SCPU or the single compute engine is to transform the data.

48. The SSD, as recited in claim 45 wherein the host interface is coupled to the host through a host interface bus.

49. The SSD, as recited in claim 48 , wherein the host interface bus is PCI Express (PCIe), Ethernet, fiber channel, infiniband, or Serial ATA (SATA).

50. A method of managing data in a solid state disk (SSD) comprising:

receiving a command and accompanying data from a host through an interface, the host being located externally to the SSD, by a storage central processing unit (SCPU), the command requiring transformation of the data;

storing the data to be transformed in a bank of memory devices coupled to the SCPU;

determining a single compute engine, coupled to the SCPU and the bank of memory devices, is to perform transformation of the data, the single compute engine being separated from the stored data in the bank of memory devices only by the interface; and

upon the determination, the single compute engine being operable to perform transformation of the stored data entirely within the SSD by exploiting the close locality of the data being transformed to the single compute engine, before the transformed data is sent to the host and to avoid the need for data movement to and from the host during the transformation,

wherein the SCPU is operable to determine whether the SCPU or the single compute engine perform the transformation of the data.

51. The method of managing data, as recited in claim 50 , further including moving the data from the bank of memory devices to a memory coupled to the SCPU when the SCPU is to perform transformation of the data.

52. The method of managing data, as recited in claim 50 , wherein the transformation of the data is searching of the data.

53. A solid state disk (SSD) comprising:

a storage central processing unit (SCPU) operable to receive from and send commands to a host located externally to the SSD, the SCPU responsive to a command requiring a search for data;

a single compute engine coupled to the SCPU; and

a bank of memory devices coupled to the SCPU and the single compute engine and configured to store the data, the single compute engine being separated from the bank of memory devices only through an interface, the SCPU operable to search for the data entirely within the SSD, before a result of the search is sent to the host such as to avoid the need for any data movement to and from the host during the search,

wherein the SCPU is operable to determine whether the SCPU or the single compute engine perform the transformation of the data.

54. The SSD, as recited in claim 53 , wherein transforming the data is searching the data.

Assignments (17)
PATENT SECURITY AGREEMENT Recorded Jun 30, 2025
From: CREELED, INC.; PENGUIN SOLUTIONS CORPORATION (DE); SMART EMBEDDED COMPUTING, INC.; SMART HIGH RELIABILITY SOLUTIONS LLC; SMART MODULAR TECHNOLOGIES, INC.; PENGUIN COMPUTING, INC.
To: JPMORGAN CHASE BANK, N.A.
Reel/Frame 071755/0001 →
RELEASE OF PATENT SECURITY INTEREST RECORDED AT R/F 058983/0001 Recorded Jun 25, 2025
From: CITIZENS BANK, N.A.
To: SMART MODULAR TECHNOLOGIES, INC.; SMART EMBEDDED COMPUTING, INC.; SMART HIGH RELIABILITY SOLUTIONS LLC; CREELED, INC.
Reel/Frame 071725/0207 →
RELEASE OF SECURITY INTEREST Recorded Feb 7, 2022
From: BANK OF AMERICA, N.A., AS AGENT
To: SMART HIGH RELIABILITY SOLUTIONS LLC
Reel/Frame 058913/0162 →
RELEASE OF SECURITY INTEREST AT REEL 043495 FRAME 0349 Recorded Feb 7, 2022
From: BARCLAYS BANK PLC
To: SMART HIGH RELIABILITY SOLUTIONS LLC
Reel/Frame 058963/0466 →
SECURITY INTEREST Recorded Feb 7, 2022
From: SMART MODULAR TECHNOLOGIES, INC.; SMART HIGH RELIABILITY SOLUTIONS, LLC; SMART EMBEDDED COMPUTING, INC.; CREELED, INC.
To: CITIZENS BANK, N.A.
Reel/Frame 058983/0001 →
SECURITY INTEREST Recorded Dec 23, 2020
From: SMART HIGH RELIABILITY SOLUTIONS LLC
To: BANK OF AMERICA, N.A., AS AGENT
Reel/Frame 054739/0064 →
SECURITY AGREEMENT Recorded Aug 9, 2017
From: SMART HIGH RELIABILITY SOLUTIONS LLC
To: BARCLAYS BANK PLC, AS ADMINISTRATIVE AGENT
Reel/Frame 043495/0349 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jul 24, 2015
From: SMART GLOBAL HOLDINGS, INC.
To: SALEEN INTERMEDIATE HOLDINGS, INC.
Reel/Frame 036175/0353 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jul 24, 2015
From: SALEEN INTERMEDIATE HOLDINGS, INC.
To: SMART WORLDWIDE HOLDINGS, INC.
Reel/Frame 036175/0368 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jul 24, 2015
From: SMART WORLDWIDE HOLDINGS, INC.
To: SMART MODULAR TECHNOLOGIES (GLOBAL), INC.
Reel/Frame 036175/0427 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jul 24, 2015
From: SMART MODULAR TECHNOLOGIES (GLOBAL), INC.
To: SMART MODULAR TECHNOLOGIES (DH), INC.
Reel/Frame 036175/0468 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jul 24, 2015
From: SMART MODULAR TECHNOLOGIES (DH), INC.
To: SMART MODULAR TECHNOLOGIES (DE), INC.
Reel/Frame 036175/0490 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jul 24, 2015
From: SMART MODULAR TECHNOLOGIES (DE), INC.
To: SMART HIGH RELIABILITY SOLUTIONS LLC
Reel/Frame 036175/0500 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jul 24, 2015
From: FASTOR SYSTEMS, INC.
To: SMART GLOBAL HOLDINGS, INC.
Reel/Frame 036175/0320 →
RELEASE OF SECURITY INTEREST Recorded Apr 3, 2015
From: SUN, DAVID
To: FASTOR SYSTEMS, INC.
Reel/Frame 035325/0914 →
SECURITY AGREEMENT Recorded Oct 27, 2014
From: FASTOR SYSTEMS, INC.
To: SUN, DAVID
Reel/Frame 034090/0063 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jan 3, 2014
From: ASWADHATI, AJOY; ASWADHATI, VIJAY
To: FASTOR SYSTEMS, INC.
Reel/Frame 031889/0919 →